Adnan KhanAdnanProduct & AI
San Francisco Bay Area
Get in touch

Source note3 min read

A Self-Evolving Agentic System for Automated Generation and Execution of Biological Protocols

Source

arXiv

arxiv.org/abs/2606.31763v1

Read the source

Source

  • Source name: arXiv
  • Source kind: paper
  • Authors / organization: Yankai Jiang, Weiting Tang, Haoran Sun, Zhenyu Tang, Yuejie Hou, Yingnan Han, Rubo Wang, Yueyuxiao Yang, Cheng Liang, Lilong Wang, Wenjie Lou, Xiaosong Wang, Lei Bai, Meng Yang
  • Venue: arXiv: cs.AI
  • Published: 2026-06-30
  • Collected: 2026-07-01
  • URL: http://arxiv.org/abs/2606.31763v1
  • PDF: https://arxiv.org/pdf/2606.31763v1
  • DOI: UNKNOWN

Plain-Language Brief

Researchers have developed a self-evolving multi-agent system called ProtoPilot for automated generation and execution of biological protocols. The system incorporates various components to ensure alignment between biological intent, quantitative procedures, device constraints, and experimental feedback. ProtoPilot was tested against 294 synthetic-biology and molecular-biology tasks derived from 98 gold-standard protocols and achieved high success rates compared to OpenTrons-AI.

What the Document Says

  • ProtoPilot is a self-evolving multi-agent system for automated generation and execution of biological protocols.
  • The system includes layer-wise verifiability, multi-agent orchestration, and a runtime-updated skill library.
  • It was tested against 294 synthetic-biology and molecular-biology tasks derived from 98 gold-standard protocols.
  • ProtoPilot achieved a Top@3 expert-preference rate of 90.2%.
  • It had an overall protocol-to-code gate pass rate of 89.5%.
  • The system also achieved an Opentrons pass rate of 88.24%.
  • This is compared with 32.35% for OpenTrons-AI.
  • Wet-lab validation produced interpretable readouts, Sanger-confirmed products, and feedback-corrected PCA-assembled DNA targets.
  • The results show that ProtoPilot can meet the execution-relevant requirements for autonomous wet-lab automation.

Key Concepts Mentioned

  • Self-evolving multi-agent system
  • Automated generation and execution of biological protocols
  • Layer-wise verifiability
  • Multi-agent orchestration
  • Runtime-updated skill library
  • Gold-standard protocols
  • Synthetic-biology tasks
  • Molecular-biology tasks
  • Expert-preference rate
  • Protocol-to-code gate pass rate
  • Opentrons pass rate
  • Wet-lab validation
  • Interpretable readouts
  • Sanger-confirmed products
  • Feedback-corrected PCA-assembled DNA targets

Claims Made by the Document

  • ProtoPilot can generate and execute biological protocols.
  • The system can meet the execution-relevant requirements for autonomous wet-lab automation.

Evidence Used

  • The description of ProtoPilot and its components.
  • The results of testing against 294 synthetic-biology and molecular-biology tasks.
  • The success rates achieved by ProtoPilot compared to OpenTrons-AI.

Methods / Approach

  • Development of a self-evolving multi-agent system (ProtoPilot) for automated generation and execution of biological protocols.
  • Testing of the system against 294 synthetic-biology and molecular-biology tasks derived from 98 gold-standard protocols.

Limitations Stated by the Authors

  • The study does not provide details on the specific challenges encountered during the development or testing of ProtoPilot.
  • It is unclear how generalizable the results are to other types of biological experiments.

Limitations Noted by the Agent

Agent interpretation — not from the source.

  • The data provided does not allow for a comprehensive evaluation of the system's performance in various real-world scenarios.
  • There is no information on the potential impact of external factors (e.g., equipment malfunctions, unexpected experimental outcomes) on ProtoPilot's performance.

Open Questions

  • What are the specific challenges encountered during the development and testing of ProtoPilot?
  • How generalizable are the results to other types of biological experiments?

Notes for Future Q&A

May include interpretation. Labeled where it does.

  • Questions regarding the potential impact of external factors (e.g., equipment malfunctions, unexpected experimental outcomes) on ProtoPilot's performance may need to be addressed through additional research or experimentation.

Raw Abstract / Excerpt

Autonomous wet-lab experimentation requires more than plausible protocol text: biological intent, quantitative procedures, device constraints and experimental feedback must remain aligned from protocol and SOP design to code and physical execution. We developed ProtoPilot, a self-evolving multi-agent system, together with an expert-grounded benchmark and evaluation framework for testing this conversion as an experimental automation problem. The framework spans 294 synthetic-biology and molecular-biology tasks derived from 98 gold-standard protocols, wet-lab expert rubrics, device-level validity gates and real experimental tests. ProtoPilot incorporates layer-wise verifiability, multi-agent orchestration and a runtime-updated skill library to generate protocols, expand SOPs, synthesize SDK-compliant code and revise workflows from wet-lab feedback. It achieved a Top@3 expert-preference rate of 90.2%, an overall protocol-to-code gate pass rate of 89.5% and an Opentrons pass rate of 88.24%, compared with 32.35% for OpenTrons-AI. Wet-lab validation produced interpretable readouts, Sanger-confirmed products and feedback-corrected PCA-assembled DNA targets, establishing a verifiable route to autonomous experimentation. Together, these results show that the evaluation framework captures execution-relevant requirements for autonomous wet-lab automation, and that ProtoPilot can meet them by converting protocol and code generation into validated execution and feedback-guided revision.